Victimization and Gender Identity in Single-Sex and Mixed-Sex Schools: Examining Contextual Variations in Pressure to Conform to Gender Norms
- 3.5k Downloads
Contextual variations in the association between gender role conformity and victimization were studied in fourth-, fifth- and sixth-grade girls (N = 469) from same- and mixed-sex schools located in lower middle class neighborhoods in two cities (Bogotá and Barranquilla) in Colombia. Self-report questionnaires were used to measure peer victimization (i.e., when a child is the target of negative actions from her/his peers) and gender identity. We hypothesized a negative relationship between typicality and peer victimization in that more gender typical children would be less victimized. We also expected to find a stronger association between gender typicality and victimization in contexts with higher levels of pressure to conform to gender norms. Results indicated that the girls in the single-sex schools endorsed significantly higher levels of gender typicality and felt pressure to conform to gender norms than the girls in the mixed-sex schools. The girls in the mixed-sex schools reported significantly higher levels of peer victimization. Girls in the same-sex schools reported lower levels of peer victimization in classrooms with high levels of perceived pressure to conform to gender norms. Multilevel modeling revealed that gender typicality was negatively associated with peer victimization among the girls in the same-sex schools and was uncorrelated with victimization in the mixed-sex school. These findings add to the database indicating that peers relationships are affected by gender and that the effects of gender are moderated by group composition.
KeywordsSingle and Mixed-sex schools Peer victimization Gender identity Middle childhood
Research on the purposes and effects of same- and mixed sex schooling can be approached from multiple perspectives (Bigler and Signorella 2011). Legitimate questions can be raised about the validity of reasons offered to justify the need for same-sex education (AAUW 1998) and one can assess the extent to which same- and mixed-sex schools achieve specific educational objectives (Hoffnung 2011), Hayes et al. 2011). Of equal importance to these concerns is an interest in the social environments that are experienced by students in these schools especially as these variations are related to gender-linked processes (Pipher 2002). In this study, a classroom-based approach was used to assess contextual differences in girls’ experiences of gender and concomitant associations with peer victimization. The purpose of the present study was to investigate the relation between two aspects of gender identity, specifically typicality and felt pressure to conform to gender norms and peer victimization in two distinct social contexts, single-sex and mixed-sex schools.
Past research on peer victimization has identified several individual level behaviours, such as aggression and withdrawal, that increase a child’s risk for victimization (Bukowski and Sippola 2001; Hodges et al. 1997, 1999). The present study was designed to extend the present literature on victimization in two critical and interrelated ways. First we wanted to go beyond the consideration of variables related to individuals per se by examining the effects of a contextual variation, specifically the differences between the experiences of girls in same-sex schools and in mixed-sex schools. Second, we wished to assess the intersection between this contextual factor (i.e., school gender composition) and an individual level variable (i.e., gender typicality). A third contextual factor that is critical to our study is its location in a cultural milieu outside of the typical locales where research on victimization has taken place (i.e., North America, Europe, and Australia). This study was conducted in Colombia in Latin America, a location that has been largely overlooked in research on gender and on victimization.
The central question of the study concerned whether the effects of gender nonconformity vary across contexts that differ in their gender composition. Extant theory and research based primarily on Western societies shows already that the frequency and predictors of victimization will likely be affected by contextual factors (Rubin et al. 2006). In this study we assess whether the gender composition of the group, as a feature of the social context, may alter the association between gender experiences and victimization. The study had three goals: (1) examine differences in gender identity and amount of victimization experienced by girls in the two types of schools: (2) assess whether gender role experience was associated with victimization; and (3) determine whether differences in victimization between the types of schools derived from differences in gender typicality and felt pressure to conform to gender norms in these schools.
Peer victimization occurs when a child is the target of negative actions from her/his peers (Kochenderfer-Ladd and Ladd 2001; Olweus 2001). Among children and adolescents these actions might include physical aggression, verbal abuse, or social manipulation (e.g. social ostracism and spreading rumors). According to Juvonen and Graham (2001) “the crucial element that distinguishes peer harassment from other types of negative encounters, such as conflict, is that there is an imbalance of power between perpetrator and target” (p.xiii). This power differential can take many forms such as differences in physical size and strength, age, or intellectual ability. Although psychologists have long been concerned with the problem of peer victimization, most research has focused on the perpetrators of hostility (i.e., bullies) rather than their targets. However, over the past decade, accumulated evidence indicates that peer victimization has a negative effect on the development and adjustment of the victim. For example, in studies conducted in Québec victimized children who were observed to be more victimized were more likely to feel lonely and depressed (Boivin et al. 1995) and to develop internalizing and externalizing problems (Hodges et al. 1999). Victimized children may interpret negative peer experiences as critical appraisals of the self, leading to internalized distress (e.g. depression, loneliness, low self-worth). Alternately, they may develop pejorative attitudes toward their peers, subsequently leading to externalizing problems such as aggression or oppositional behaviour (Prinstein et al. 2001). As such, victimization can be an important determinant of a child’s development and well-being and, therefore, deserves our attention as psychologists, educators and health practitioners. The present study focuses on victims of peer harassment, rather than on aggressors, and examines links between victimization and gendered experiences.
The present study is unique in its use of a multilevel approach that assesses contextual variations in processes and effects at the level of the individual. Bukowski and Sippola (2001) propose a dynamic model of victimization wherein a balance is struck between individual characteristics and group needs, defined as cohesion, homogeneity, and evolution. Group dynamics are oriented towards achieving group goals and persons who facilitate these goals are given special privileges (popular children) while persons who impede these goals are treated in ways that minimize their participation in the group (victimized children). According to this view, victimization is a process through which an individual is denied access to a group because he or she is seen as a threat to the attainment of the group’s goals. In so far as gender development is a process of becoming a member of one of two gender groups, children who are perceived as gender nonconforming may be seen as a threat to group cohesion and homogeneity. As we show in the next section, key findings from research on gender point to the importance of gender as a factor underlying victimization.
It is known that childhood and adolescent peer groups function as influential contexts for gender socialization (Harris 1995; Leaper and Friedman 2007). Peers have been shown to be vigilant in their enforcement of gender norms and generally disapproving of cross-gender-typed behaviour (Martin 1989). It has been proposed that gender atypical behavior is likely to incur more sanctions and other negative reactions from peers (Crick and Dodge 1994). This process may account for the relatively greater adjustment problems seen in children who engage in gender atypical behaviour (Crick 1997; Henington et al. 1998). Furthermore, researchers have identified difference (e.g., being overweight, having a disability, or poor school performance) as a vulnerability factor for victimization in adolescence (Young and Sweeting 2004). Together these studies, typically conducted with American samples, show that children who do not conform to group norms of gender expression may be at increased risk for victimization by their peers. This emphasis on typicality is already apparent in the explicit concern among clinical psychologists that children and adolescents whose behaviour is deemed atypical enough to warrant a diagnosis of gender identity disorder are at greater risk for social rejection and psychological disorders (Cohen-Kettenis et al. 2003; Savin-Williams 1994). We are interested in examining whether gender atypical behaviour in the subclinical range is also associated with peer relationship problems and psychological distress and whether these effects vary as a function of the gender composition of the school context.
An individual’s sense of being either a male or a female person is thought to be a core element in the developing sense of self. The emergence of a gender identity and growing understanding of the stability of social group membership affects children’s motivation to learn about gender, to gather information about their gender group, and to act like other group members (Ruble and Martin 1998). In other words, children’s recognition of the social significance of gender motivates them to learn about and comply with gender norms.
Recent research on gender identity development (Egan and Perry 2001) conducted in the United States suggests a developmental conceptualization along four dimensions: inter-group bias (how much one favours their own gender category over the other), contentedness (how content one is with socially prescribed gender), felt pressure (how much pressure one feels to conform to gender norms) and typicality (how typical one feels for one’s gender). Prior to middle childhood, children strive to comprehend their membership in a gender category (gender constancy), exhibit considerable inter-group bias, experience strong pressure to conform to gender stereotypes, and generally report high satisfaction with their gender assignment (Ruble and Martin 1998). As such, at this age, these components of gender identity are positively associated with adjustment outcomes such as self-esteem (Yunger et al. 2004). However, by middle childhood, most children have learned to temper certain developmentally immature components of gender identity – namely felt pressure to conform to gender norms and inter-group bias and as a result exhibit lower levels of these behaviours (Ruble and Martin 1998). A possible explanation for this attenuation is that throughout childhood increasing emphasis is placed on autonomy and individuality such that feeling strong pressure to conform to gender norms in middle childhood would become negatively associated with children’s well-being. Indeed, a small but significant body of research, typically conducted with American children, has confirmed this association (Carver et al. 2003; Egan and Perry 2001; Yunger et al. 2004).
Cognitive development during middle childhood, such as improved social comparison skills and the ability to infer stable, abstracted attributes of the self set the stage for the emergence of an additional component of gender identity namely typicality or how typical one feels for one’s gender (Yunger et al. 2004). Research examining the relationship between gender typicality and self-worth has revealed a positive direct association, such that within individuals, high gender typicality was associated with higher self-worth and low gender typicality was associated with lower self-worth (Carver et al. 2003; Egan and Perry 2001; Yunger et al. 2004). Recently, Smith and Leaper (2005) used a sample from California to test an implicit assumption of this finding: gender typical children have higher self-esteem because their peers accept them, and gender atypical children have lower self-esteem because their peers do not accept them. In other words, the authors wished to address the role of the peer group in the social process of gender development. They found that the positive association between gender typicality and perceived self-worth was mediated by a measure of perceived peer acceptance. Put another way, gender atypical adolescents had comparable self-worth to gender typical adolescents as long as they were accepted by their peers. This research speaks to the importance of considering the peer group when studying gender development.
These previous findings tell us that in middle childhood (a) it is important to be a typical girl/boy and (b) it is important not to feel strong pressure to conform to gender norms. If viewed from a contextual framework, one can expect that the association between typicality and well-being will be moderated by the degree of felt pressure in the environment. Within the rubric of the current study, victimization is our index of well-being and we are testing whether the association between gender typicality and victimization differs as a function of the level of felt pressure to conform to gender norms (a peer level process) within the two types of school.
Importance of Context – Gender Composition of the School
Although children first learn social and emotional skills within their families, it is the school setting that provides the first significant experience for most children with respect to negotiating social roles, expectations, hierarchies, and conflicts in larger groups (Merrell et al. 2006). A premise of the present study is that peer group dynamics will vary as a function of its gender composition. One can expect that the impact of gender composition will be especially strong for gender-related concepts and processes. Research on dyadic interactions among children drawn from communities in North America has shown already that the gender composition of the dyad affects the gender typicality of children’s behavior (Miller et al. 1986; Serbin et al. 1982). For example, in a study of social interaction, Maccoby and Jacklin (1987) found girls to be more passive than boys during cross-gender interactions. Interestingly, the gender difference could only be seen when the gender of the interactant partner was considered: girls were not more passive than boys overall, but rather were only more passive when observed during their interactions with boys (Maccoby and Jacklin 1987).
A logical extension of this research could be an examination of gender effects within group processes. In fact, current research and theorizing from an American context considers “how gender is implicated in the formation, interaction processes, and socialization functions of childhood social groupings” (Leaper and Friedman 2007, p. 563). Children learn how to behave outside the home by becoming members of, and identifying with, a social group – typically the same-gender peer group (Rubin et al. 2006). With the acquisition of a gender self-concept, children form a social identity of themselves as a member of a particular gender group (Leaper 2012). Being a member of a group typically leads to in-group bias (Sippola et al. 1997). Children value their in-group membership and become sensitive to how others view them in order to gauge their status within the group (Bukowski et al. 1993). In this manner, same-gender peer groups, like all groups, tend to promote within-group and between-group processes.
Harris (1995) applies intergroup processes (Sherif et al. 1961; Tajfel 1982) to gender socialization processes. Accordingly, children are expected to behave in gender-typed ways most consistently when gender-segregation is strong and when same-gender in-groups and opposite-gender out-groups are formed. Bigler and Liben (2007) have proposed a theoretical framework, called developmental intergroup theory, to explain the formation of social stereotypes and prejudice. A portion of the experimental evidence in support of the theory suggests that higher intergroup bias results from explicitly labeling groups, such as beginning the day by stating “good morning boys and girls” (Bigler et al. 1997, 2001). Perceptually marking group membership, such as assigning red or blue t-shirts, (Bigler 1995) has also been found to increase intergroup bias. This experimental manipulation is thought to act as a proxy for other psychologically salient characteristics such as gender, race, age and attractiveness. In light of intergroup theory, we are interested in examining whether the presence or absence of boys, in other words of another salient social group, alters the association between gender typicality and victimization.
Most studies addressing differences between single and mixed-sex schools have been conducted in high schools with adolescent samples. Despite age differences and concomitant developmental disparities, findings from this research remain relevant to the current project. For example, Lee and Marks (1990), found that girls educated in Catholic single-sex in the United States compared with girls from mixed-sex schools endorse less stereotypical views of gender roles into their college years. Likewise, Lee and Bryk (1986) found that American girls in single-sex schools had less stereotypical adult gender role attitudes and had higher self-esteem. In another study, Stables (1990) found greater gender polarization of attitudes towards school subjects (particularly physics) in mixed-sex schools. Taken together these studies suggests that for adolescent girls, the presence of boys fosters the development of more gender-typed attitudes and adherence to gender role norms.
The Current Study
The study was conducted in two cities in Colombia in Latin America. There are at least four advantages to conducting a study of this sort in Colombia. One advantage derives from the relative strength of gender-based expectation in Latin America. Although there is a specific lack of immediately relevant research on gender roles conducted with Colombian samples, ideas presented in recent review papers regarding the place of gender roles within the Latin American context suggest that gender is an especially powerful social construct (Chant 2003; Hoffman and Centeno 2003). Two conclusions are especially pronounced in the available research on gender roles in Latin America. They are (a) that the significance of gender roles is stronger in Latin American contexts than it is in North America and Europe, and (b) that the more restrictive features of gender role expectations for women in Latin America derive from a stronger emphasis on machismo among Latin American men compared with men from North America and Europe (Diekman et al. 2005). This latter point is especially relevant for the current study’s comparison of girls from all-girl schools with girls from mixed-sex schools. The claim that gender roles are emphasized heavily in Latin America reinforces our expectation that violations of gender role expectations will be sanctioned within the peer group. The idea that the emphasis on a restrictive and traditional role for women comes from the emphasis on machismo among men, suggests that the pressure to conform to the female gender role may be stronger in a mixed-sex school context than in an all-girl context.
A second advantage of conducting the study in two cities in Colombia comes from the cultural heterogeneity within Colombia. Like many other countries, Colombia is a political entity rather than a homogeneous cultural context. There is quite a bit of cultural variations between the regions within Colombia. Because the two cities where the study was conducted are within the same national political context, the schools in these cities were subject to the same government based regulations regarding schooling practices and curriculum. Nevertheless these two communities vary considerably in the features of their cultural contexts. Whereas Barranquilla has a Caribbean culture that emphasizes public displays of happy emotions and an openness to new experiences, Bogota has an Andean culture characterized by stoicism and self- restraint. In Barranquilla there is an emphasis on outdoor living due to the year-round high temperatures; in Bogota, which is closer to the equator but is situated at a very high altitude – nearly 2,800 m above sea level - the temperature is much milder and most functions go on indoors. In contrast to Barranquilla, public displays of the body are kept to a minimum in Bogota and persons use a more formal mode of dress. Due to their location in the same country but with different cultural traditions, these two cities present continuity in the structural features of schools but variability in cultural factors. Although we were not interested in the effects of this cultural variability per se it does offer us the opportunity to assess whether any observed differences between girls in all-girl and mixed-sex schools would vary across two different contexts.
A third is related to the general frequency of all-girl schools in Colombia. Whereas all girls schools in North America are often private institutions that typically serve upper-middle families or are “pilot” projects linked to social policy goals that derive from ideas about the consequences of gender-related processes within schools (Bigler and Signorella 2011), all girls schools in Colombia remain available to families from different SES backgrounds. Although all-boys schools are increasingly less frequent in Colombia, all-girls schools are still a normal component of the educational offering for girls. Accordingly, when one studies the girls in all-girls schools one is not likely to be examining a unique subset of the overall population. Instead one is studying a group of girls who are likely to be largely representative of the general population of girls.
Related to this third advantage is the Colombia’s status as part of the “majority” world. Although there is a general recognition of the importance of studying individuals from “majority” (i.e., non-Western) societies (Arnett 2008), research outside of well-developed countries remains rare. For example, all of the studies in the first volume in this special set of papers in Sex Roles on same-sex education were conducted in countries in the developed world. The general lack of research on social developmental processes with samples from developing countries is problematic as the findings from “non-majority” contexts may not produce findings that can be generalized to the vast majority of world’s population. It is known already that there are differences between children from WEIRD (Western, Educated, Industrialized, Rich, and Democratic) and non-WEIRD cultural contexts (Henrich et al. 2010a, b). For this reason, one needs to be concerned about the applicability of findings observed with American and European samples to samples in Latin America. Accordingly, a fourth advantage of studying lower middle class girls from Colombia is that they may more closely approximate girls from the rest of the world, at least in terms of SES, than girls from North America.
Hypothesis 1: Contextual and Grade Differences in Gender Identity & Peer Victimization
Past research has shown higher levels of both typicality and felt pressure to conform to gender norms in single-sex schools as compared to mixed-sex schools (Drury and Bukowski 2009). We expected to find this again. We are not aware of any research or theory addressing differences in levels of victimization in single-sex and mixed-sex schools; as such this question was exploratory. Based on past research (Egan and Perry 2001), we expected mean levels of felt pressure to conform to gender norms to decrease across grade. Finally, based on the Gender Intensification Hypothesis proposed by Hill and Lynch (1983), we expected increases in mean levels of typicality across grade.
Hypothesis 2: Association between Gender Identity and Peer Victimization
Research examining the associations between gender identity and adjustment suggests a positive association between gender typicality, social competence, and well-being (Egan and Perry 2001; Smith and Leaper 2005). Based on this as well as on another study with adolescents indicating an association between gender atypical behaviour and problems with peers and psychological distress (Young and Sweeting 2004), we hypothesized a negative relationship between typicality and peer victimization in that more gender typical children would be less victimized. Grade differences were exploratory.
Hypothesis 3: Contextual Differences in Felt Pressure
We are not aware of past research examining gender typicality and victimization in single-sex and mixed-sex schools; as such this portion of the study was exploratory. Based on research suggesting that gender atypical behaviour is sanctioned by peers (Crick 1997), we expected to find a stronger association between gender typicality and victimization in contexts with higher levels of pressure to conform to gender norms.
The participants were 469 fourth, fifth and sixth grade girls (Mean for age = 10.13 years, SD = 1.22) attending either single-sex schools (n = 327) or mixed-sex schools (girls n = 142). The schools were located in two Colombian cities, one (Bogotá) located in the Andean region of the country and the other (Barranquilla) located on the Caribbean coast, with one single-sex and one mixed-sex school in each city. All four of the schools would be labelled as private within the Colombian system, meaning that the government does not own them. However, unlike the North American system, the public/private distinction is not confounded by socio-economic status in that children can attend private schools for free. As is the case for many schools in Colombia, religious orders are responsible for the administration of the school, however, not necessarily the pedagogy. All of the children in the study wore uniforms to school. The schools were of comparable size, located in lower socio-economic neighbourhoods. Based on school records and parent reports, the majority of children were from lower-middle class backgrounds.
It is important to note that that the selection of a school for a child often reflects a complex set of considerations. These factors may include practical considerations, such as the school’s location or the availability of other schools, as well as parental perceptions of the sort of environment they would prefer for their child. It would not be possible to claim that the decision for a child to go to one school instead of another is due to a single factor. In Colombia, children typically attend a school in which the other students are drawn from the same neighbourhood and, as result, are typically from the same sector of the SES spectrum. Children are required to attend schools that fall within their neighbourhood SES spectrum.
Multidimensional Gender Identity Inventory (Egan and Perry 2001)
The Gender Identity Inventory is a 30-item self-report measure. For the purpose of this study a revised 17-item version was administered. This version includes three subscales, two of which were used in the current study. Gender typicality (α = .70), a two-item variable (e.g., “I like to do the things that most girls like to do), and felt pressure to conform to gender norms (α = .73), a four-item variable (e.g., “It would bother the kids in my class if I acted like a boy”). These items aimed to measure a group characteristic and therefore were aggregated at the peer group level. Scores on both gender identity variables represent a mean of the items. Items are rated on a 5-point likert scale (1 = Really disagree, 5 = Really agree).
Self-Assessment: What am I Like?
The self-assessment is a 25-item measure tapping how children see themselves across several dimensions, which included 2 victimization items (α = .70) (e.g., “others call me bad names” and “others treat me badly”). Each participant’s score on victimization is a mean of the two items. Items are rated on a 5-point likert scale (1 = Really disagree, 5 = Really agree).
Permission for the study was first obtained from the school principals. The researchers then described the purpose of the study, the time commitment, confidentiality, and distributed a consent form to be signed by the students and their parent/guardian. Using this recruitment procedure, a participation rate of approximately 89% was obtained.
Items used to measure gender typicality, felt pressure to conform, and peer victimization
I am just like the girls in my class.
Soy como las otras niñas de mi curso.
I think that I am very similar to the other girls in my class.
Pienso que soy muy parecido a las otras niñas de mi curso.
Felt Pressure to Conform
It bothers the kids in my class when a girl acts like a boy.
A mis compañeros les molesta cuando una niña actúa como un niño.
It would bother the kids in my class if I acted like a boy.
Mis compañeros se molestarían si yo actuara como niño.
Kids in my class don’t like girls who act like boys.
A mis compañeros no les gustan las niñas que actúan como niños.
Most kids would think it is weird if I did something that boys like to do.
A la mayoría de mis compañeros les parecería raro si yo hiciera algo que usualmente hacen los niños.
Others call me bad names.
Los otros me ponen apodos ofensivos.
Others treat me badly.
Los otros me tratan mal.
Bivariate associations between variables for participants in Grade 4, 5 & 6
1. Gender Typicality
2. Felt Pressure
1. Gender Typicality
2. Felt Pressure
1. Gender Typicality
2. Felt Pressure
Hypothesis 1: Group Differences in Gender Identity & Levels of Peer Victimization
Mean differences in victimization, typicality and felt pressure between single and mixed-sex schools and between grades were assessed with a repeated measures analysis of variance. It had been hypothesized that felt pressure to conform would decrease with age whereas typicality was expected to increase. Levels of typicality and felt pressure were expected to be higher in the single-sex schools than in the mixed-sex schools. In this analysis the grade and type of school were used as the between subjects factors and “type of scale” (i. e., typicality, pressure, and victimization) was used as a within subjects factor. Three significant effects were observed. A univariate effect was observed for the type of scale variable, F(2, 910) = 81.86, p < .001; this main effect was qualified by significant interactions between type of scale and grade, F(4, 910) = 2.84, p < .02 and between type of scale and type of school, F(2, 910) = 31.04, p < .001.
The two interactions were clarified with simple effects tests conducted with each of the three scales (i.e., typicality, pressure, and victimization). When typicality was used as the dependent measure significant effects were observed for the type of school variable, F(1, 910) = 3.502, p < .05 and for grade, F(1, 910) = 4.84, p < .005. Scores on typicality were higher in the all girls schools than in the mixed sex schools (Ms = 2.80 (sd = 1.30) and 2.51 (sd = 1.22)), and they decreased from grade four to grades six (Ms and sds) = 2.97 (1.27), 2.71 (1.28), and 2.38 (1.23) for grades four, five and six respectively). Only one effect was observed when the pressure score was used as the dependent variable. The significant effect for type of school, F(1, 910) = 3.7.78, p < .001 was associated with substantially higher scores in the all girls schools than in the mixed sex school (Ms = 3.71 (sd = 1.05) and 3.00 (sd = .88)). An effect for type of school was also the only effect observed with the victimization score, F(1, 910) = 23.96, p < .001. Higher scores were observed among the girls in the mixed-sex school than among the girls in the all girls school (Ms and sds) = 2.52 (1,28) and 1.95 (1.14).
Hypothesis 2: Association between Gender Typicality and Peer Victimization
Multilevel modeling, conducted with HLM 6.06 (Bryk and Raudenbush 1992), was conducted to examine the association between gender typicality and victimization, and to test variations in these association across peer groups. We expected to find a negative relationship between the constructs in that higher levels of typicality would be associated with lower levels of victimization.
First, an unconditional model, which included only victimization as an outcome, was run to estimate the variability between and within groups. Based on intra-class correlations 4.0% of the variability was between groups and 96.0% of the variability was within groups.
Next, typicality was included as a predictor of peer victimization at level 1. Results showed that gender typicality (B = -.14, t = -2.41, p < .05) was a significant predictor of victimization. The negative slope confirmed our hypothesis that children who reported more typicality reported less victimization. Based on the reduction of sigma-squared typicality explained 5% of the within group variance. Finally, it was noted that the slope the association between typicality and victimization was random (χ2(19) = 39.53, p < .05) indicating that it varied across groups. This latter finding is evidence that gender typicality is differentially associated with victimization across classrooms.
Hypothesis 3a: Contextual Variations as a Function of Type of School
Hypothesis 3b: Contextual Differences in Felt Pressure to Conform
The present study provides an important new look at the processes related to peer victimization during the older school-age years. Whereas the lion’s share of the studies in the literature on peer victimization have focused on specifically individual-level behavioral measures such as aggression and withdrawal, we chose to study it from a gender-based contextual perspective. Furthermore, we hoped to galvanize research initiatives in “majority” (i.e., non-Western) societies. In general, our aim was to further understanding of peer victimization by examining its relationship to gender, specifically gender identity and felt pressure to conform to gender norms. We know from past research that children who have been diagnosed with gender identity disorder are frequently targets of peer victimization (Cohen-Kettenis, et al. 2003). We hoped to extend this finding to a subclinical population of children in order to better understand the role of gender in children’s social networks. In addition, in light of the importance of considering context when studying social processes, we examined the associations between the variables in single-sex and mixed-sex schools, expecting to find differences between the environments. Keeping in mind the complexity of the issues at hand, findings from the current study certainly further our understanding of gender-based victimization and have implications for the single-sex/mixed-sex schooling debate.
Hypothesis 1: Interpretations & Implications
Although we are advocating for an exploration of social processes extending beyond the limits of mean differences between groups, we nonetheless believe that the information conveyed by mean differences between contexts reflects important process-oriented considerations. When comparing the girls in the single-sex schools with the girls in the mixed-sex schools, two mean differences in the variables of interest appeared. First, the girls in the single-sex schools endorsed significantly higher levels of gender typicality and felt pressure to conform to gender norms than the girls in the mixed-sex schools. In other words, we have shown that differences in levels of girls’ gender-typed behaviours are associated with the gender composition of the school context. Second, the girls in the mixed-sex schools reported significantly higher levels of victimization.
The gender identity findings suggest that a single-sex environment may produce a social dosage effect (Martin and Fabes 2001): spending more time with same-gender peers leads to increased feelings of pressure to conform to gender norms and higher levels of gender typicality. This could be understood as a narrowing of experience or, conversely, as an expansion of possible experiences. On the one hand, an environment characterized by high levels of pressure to conform to gender norms may socialize higher levels of gender typicality. On the other hand, past research has shown that both males and females in single-sex schools engage in more cross-gendered behaviour (Stables 1990), which could provide an alternate interpretation. Engaging in cross-gendered behaviour may serve to expand the parameters of gender roles to include more behaviours and as a result more individuals in single-sex schools report being gender typical in conjunction with reporting higher levels of pressure.
The girls in the mixed-sex environment rated themselves as less typical and as feeling less pressure to conform to gender norms. Perhaps the presence of boys leads to a narrowing of the category “typical girl” resulting in less girls reporting being gender typical. It could be that in general, girls in this context feel less pressure to conform to gender norms because the norms are so stringent, and thus perceived as unattainable. However, based on Bigler and Liben developmental intergroup theory (2007), perhaps if we were to isolate the girls in this context who do report being gender typical, they would report higher levels of pressure to conform, as a form of intergroup bias, as compared to girls in single-sex schools. This would be an interesting avenue for future research. If we extend this perspective to the process of victimization, perhaps there is more victimization in the mixed-sex schools because the narrower definition of “girl” excludes more individuals, thus leaving them vulnerable to peer victimization. Alternately, the presence of boys may contribute to higher levels of victimization in mixed-sex schools. In this context, cross-gender victimization could take the form of sexual harassment. To our knowledge, research has yet to determine whether gender nonconformity or gender conformity increases an individual’s risk for sexual harassment. As such this would also be an important direction for future research.
Hypothesis 2: Interpretations & Implications
With regards to the association between gender typicality and victimization, our hypothesis was confirmed in that the association was negative. This finding is in accordance with past research with adolescents (Young and Sweeting 2004) as well as research with clinical samples of children and adolescents diagnosed with gender identity disorder (Cohen-Kettenis et al. 2003). The repercussions of this finding are clear: if these processes are already evident in children in middle childhood then our efforts to prevent them should be focused on this age group, and possibly younger children. Future research is needed to address whether these processes begin at an early developmental period than the one studied here.
Hypothesis 3a: Interpretations & Implications
With regard to contextual differences in the association between typicality and victimization, our results indicated that gender typicality might protect against victimization, however only in single-sex schools. For girls in this context, whether or not you consider yourself to be a typical girl has implications for your social functioning. This finding highlights the social importance of gender typicality in this context; being a typical girl protects you from victimization, and conversely, being atypical places you at risk for peer victimization.
For the girls attending mixed-sex schools, we found no significant association between gender typicality and peer victimization. Considering that mean levels of peer victimization were higher in this context, we might conclude that girls in this context are being victimized for reasons other than whether or not they consider themselves to be a typical girl. Therefore, the abovementioned suggestion that higher levels of peer victimization in mixed-sex schools may be associated with higher levels of gender atypical behaviour is in fact not supported by our data.
Hypothesis 3b: Interpretations & Implications
As expected, the negative association between typicality and peer victimization was stronger in contexts characterized by high levels of pressure to conform to gender norms. In other words, gender atypical children report more peer victimization in environments where children feel a lot of pressure to conform to gender norms. In the present study, single-sex schools represent such an environment, however, our findings suggest that it is perhaps not the type of school that contributes to the link between gender typicality and victimization but rather the levels of felt pressure to conform. As such, we might expect a strong association between gender atypicality and peer victimization in mixed-sex schools characterized by high levels of pressure to conform to gender norms.
Strengths and Limitations
The major strength of this study is its assessment of the role of gender-based processes to explain individual differences in social experience and in two different social contexts. To our knowledge this study is the first examination of the role of gender-related variables in the processes associated with peer victimization. Of course, a particular strength of the study is the observation that gender plays a very different role in girls’ lives depending upon whether boys are also present. This latter finding makes this study important for our understanding of the intersection between gender, peer relations, and educational policy. The focus on within-gender comparisons was another strength of the study. Gender research is often too focused on between-gender comparisons and as a result overlooks the substantial within-gender differences.
Like other studies, this project has some methodological limitations. First, it was difficult to identify selection criteria of girls in single-sex schools, as such it was assumed that the girls were distributed randomly between the types of schools and that there were no pre-exiting differences between the groups. These possible systemic and yet unidentified differences between the girls in the single-sex schools and the girls in the mixed-sex schools might be confounded in our analyses of between-group processes. Likewise, we did not identify any school variables that could have been controlled for in our analyses.
Second, there is the problem of directionality in that without a longitudinal design it is impossible to determine whether gender typicality and felt pressure contribute to victimization or vice versa. Finally, though the fact that the study was conducted in Colombia is one of its strengths, having a Colombian sample without a comparable Canadian sample makes it difficult to interpret findings in light of past research conducted in a North American context. As such, we find ourselves without cultural-specific empirical evidence to guide our interpretations. What follows then are speculations regarding the significance of our findings within the Colombian cultural context. In so far as generalizations can be made about a culture, Colombian culture is characterized by catholic values that dictate relatively traditional gender roles. We speculate that the adherence to more traditional gender roles may have several implications for the current project. The following are two examples, which naturally represent a non-exhaustive exploration of possible implications. Firstly, adherence to more traditional gender roles may have amplified previous findings (Egan and Perry 2001), suggesting that the measured phenomenon is positively associated with traditionalism such that more traditional societies would yield a stronger effect and less traditional societies a weaker one. Or, it may have acted to dampen the effect, possibly as a result of lower awareness of gender roles in the general population. In this sense, more traditional societies presumably engage in less deconstruction, analysis and dialogue about gender roles and therefore have a less articulated understanding of the concepts. If gender roles are taken for granted and thereby not questioned, then responses to queries about them may be more superficial causing it to be harder to detect an effect.
This study was unique given that, to date, little is known about gender-based victimization processes in childhood. We are proposing that gender-related social processes can have both positive and negative effects on an individual’s well-being and adjustment. Given the intensification of gender-typed behaviours (and possibly concomitant adjustment problems) that are seen at the transition to adolescence (Hill and Lynch 1983), we aimed to extend our understanding of these social processes by studying them in the developmental period preceding adolescence. Furthermore, we believe this research has implications for sexual minority youth in that these individuals are often targets of harassment based on perceived lower levels of gender conformity, which their aggressors equate with evidence of homosexuality or bisexuality. Considering the severity of the harassment faced by sexual minority youth in adolescence, as developmentalists, we feel strongly that these processes should be studied as they emerge in childhood in order to better design prevention programs. With regards to the our contribution to the single-sex versus mixed-sex schooling debate, the vast majority of research addressing differences between the types of schools has been conducted with high school and college populations. The current study extends this research into the earlier years of elementary school.
Implications for Single vs. Mixed-Sex Schooling Debate
Although the present study was not conducted as a comprehensive comparison of the relative advantages or drawbacks of single and mixed-sex schools, the findings do provide information about the social experiences and the social dynamics in these school environments. As one might expect when one is dealing with complex contexts, the findings present a textured image of the social environment within these schools. One aspect of the findings implies that all-girl schools have an advantage over mixed- sex schools. Indeed the mean level of peer victimization in the all-girl school is significantly lower than that of the mixed-sex school. This finding shows that levels of felt victimization are lower for girls in an environment that does not include boys. At the same time, other findings appear to favour mixed-sex schools over all-girl schools. Specifically, the pressure to conform to gender roles and the negative repercussions of not conforming are stronger in all-girl schools relative to the mixed-sex schools. In the absence of boys the “social dosage” of gender roles is higher as are the negative effects of non-conformity. Taken together these findings fail to show that one of these environments in necessarily better than the other.
This study confirms the inextricable inter-relatedness between gender and peer relations. The social experiences of the girls we studied are “gendered” in two important ways. Girls in all-girl and mixed-sex schools differ from each other in the amount of peer victimization they experience and they differ in the processes that appear to account for individual differences in peer victimization. Overall the present findings support a social dosage model of gender effects. Pressure to conform was stronger in all-girl contexts and the effects of atypicality were stronger in these contexts as well. These findings add to the database indicating that peers relationships are affected by gender and that the effects of gender are moderated by group composition.
Work on this paper was supported by (a) grants from the Social Sciences and Humanities Research Council of Canada (SSHRC) and the Fonds Québécois de la Recherche sur la Société et la Culture (FQRSC), (b) fellowships awarded to the first and third author from the SSHRC and the FQRSC, and (c) a Concordia University Research Chair awarded to the second author.
- AAUW. (1998). Gender gaps: Where schools still fail our children. Washington, DC: American Association of University Women Educational Foundation.Google Scholar
- Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical linear models: Applications and data analysis methods. Newbury Park, CA: Sage.Google Scholar
- Bukowski, W. M., & Sippola, L. K. (2001). Groups, individuals, and victimization: A view of the peer system. In J. Juvonen & S. Graham (Eds.), Peer harassment in school: The plight of the vulnerable and victimized (pp. 355–377). New York: Guilford Press.Google Scholar
- Chant, S. H. (2003). Gender in Latin America. New Brunswick, NJ: Rutgers.Google Scholar
- Cohen-Kettenis, P., Owen, A., Kaijser, V. G., Bradley, S. J., & Zucker, K. J. (2003). Demographic characteristics, social competence, and behavior problems in children with gender identity disorder: A cross-national, cross-clinic comparative analysis. Journal of Abnormal Child Psychology, 31, 41–53. doi:10.1023/A:1021769215342.PubMedCrossRefGoogle Scholar
- Drury, K., & Bukowski, W. M. (2009). Gender identity and well-being in early adolescence: Exploring the roles of the peer group and the gender composition of the school context. Unpublished Manuscript.Google Scholar
- Juvonen, J. (Ed.), & Graham, S. (Ed.). (2001). Peer harassment in school: The plight of the vulnerable and victimized. New York, NY US: Guilford PressGoogle Scholar
- Kochenderfer-Ladd, B., & Ladd, G. (2001). Variations in peer victimization. Relations to children’s maladjustment. In J. Juvonen & S. Graham (Eds.), Peer harassment in school: The plight of the vulnerable and victimized (pp. 25–48). New York: The Guilford Press.Google Scholar
- Leaper, C. (2012). Gender development during childhood. In P. D. Zelazo (Ed.), Oxford handbook of developmental psychology. New York: Oxford University Press.Google Scholar
- Leaper, C., & Friedman, C. K. (2007). The socialization of gender. In J. E. Grusec & P. D. Hastings (Eds.), Handbook of socialization: Theory and research (pp. 561–587). New York: Guilford Press.Google Scholar
- Olweus, D. (2001). Peer harassment: A critical analysis and some important issues. In J. Juvonen & S. Graham (Eds.), Peer harassment in school: The plight of the vulnerable and victimized (pp. 3–20). New York: The Guilford Press.Google Scholar
- Pipher, M. (2002). Reviving Ophelia: Saving the selves of adolescent girls. New York: Balantine.Google Scholar
- Rubin, K.H., Bukowski, W., & Parker, J.G. (2006). Peer interactions, relationships, and groups. In W. Damon (Series Ed.) and N. Eisenberg, (Vol. Ed), Handbook of child psychology: Vol. 3: Social, emotional, and personality development (pp. 571-645). New York: Wiley.Google Scholar
- Ruble, D. N., & Martin, C. L. (1998). Gender Development. In W. Damon (Ed.), Handbook of Child Psychology (pp. 933–1016). New York: Wiley.Google Scholar
- Savin-Williams, R. C. (1994). Verbal and physical abuse as stressors in the lives of lesbian, gay male, and bisexual youths: Associations with school problems, running away, substance abuse, prostitution, and suicide. Journal Of Consulting And Clinical Psychology, 62(2), 261–269. doi:10.1037/0022-006X.62.2.261.PubMedCrossRefGoogle Scholar
- Sherif, M., Harvey, O. J., White, B. J., Hood, W. R., & Sherif, C. W. (1961). Intergroup cooperation and competition: The Robbers Cave experiment. Norman, OK: University Book Exchange.Google Scholar
- Sippola, L., Bukowski, W. M., & Noll, R. B. (1997). Age differences in children’s and early adolescents’ liking for same-sex and other-sex peers. Merrill-Palmer Quarterly, 43, 547–561.Google Scholar